The EEF rates metacognition and self-regulation as one of the highest-impact strategies available to schools — +7 months additional progress. Every school improvement plan in Scotland mentions it. Very few can show evidence that it is happening in classrooms.

This guide walks through what an improvement cycle looks like when it is built on structured observation evidence rather than assumption. The school is hypothetical. The data patterns are not — they are drawn from what observation evidence consistently reveals about metacognition in practice.

Term 1: The baseline

What the school thought was happening

Hillpark Academy's improvement plan listed "embed metacognition across all departments" as a priority for the third consecutive year. Staff had attended a twilight session on the EEF guidance report in 2024. Posters about metacognitive strategies were visible in most classrooms. Its self-evaluation for learning, teaching and assessment rated the school as Good, with metacognition cited as an area of strength.

What the observation evidence showed

The school ran 40 learning visits across the first half-term using the full taxonomy — no specific lens, just the standard observation framework capturing all 59 practices. The Practice Profile view in the analytics showed this:

Practice Profile — Metacognition & Self-Regulation (Term 1, 40 visits)

Metacognition practices appeared in 18 of 40 visits (45%). But the qualifier breakdown told a sharper story:

Qualifier breakdown — Metacognition practices (Term 1)

The headline number — metacognition in 45% of visits — looked reasonable. The qualifiers revealed the reality. Of 23 total metacognition taps, 15 described weak or absent practice: thinking was implicit rather than modelled, reflection was surface-level rather than structured, goals were vague, strategy instruction was absent, and monitoring was teacher-directed rather than pupil-led.

The department breakdown added a further dimension:

Department comparison — Metacognition practices (Term 1)

No department was embedding metacognition systematically. English had the highest frequency, but even there, no explicit strategy instruction was observed. The school's self-evaluation of "Good" for metacognition was not supported by the evidence.

The evaluation conversation

This data was presented to the senior leadership team not as a criticism but as an evidence picture. Three findings drove the conversation:

First, metacognition was happening at surface level — "What did you learn today?" at the end of lessons — rather than as structured strategy instruction embedded throughout. The EEF guidance is explicit: metacognition has to be taught, not assumed. Asking pupils to reflect without teaching them how to reflect is not metacognition.

Second, strategy instruction was virtually absent. Across 40 visits, not one observation recorded explicit teaching of a named learning strategy. Teachers were modelling thinking processes occasionally, but they were not naming the strategy, explaining when to use it, or giving pupils the opportunity to practise it. The "strategy modelled then practised" qualifier was tagged zero times.

Third, the department variation showed this wasn't a pockets-of-weakness problem. It was school-wide. The twilight session from 2024 had not translated into practice.

This is the point where the data connects to HGIOS4. Quality Indicator 2.3 asks: "How good is the quality of teaching?" The evidence sub-theme 2.3.2 (Quality of teaching) expects to see "well-judged use of a range of teaching approaches including direct teaching, investigation and inquiry, and developing critical and creative thinking." The observation evidence showed that metacognitive teaching approaches were not part of the school's instructional repertoire in any consistent way. The self-evaluation needed to be revised — from Good to Satisfactory for this specific dimension, with an evidence-based rationale for the change.

Term 2: The intervention

What changed

The school took three actions based on the evidence:

1. A Metacognition observation lens was created. Instead of asking observers to look at everything, learning visits in Term 2 used a focused lens highlighting the five metacognition practices. This meant observers knew exactly what to look for, received a pre-visit briefing on what metacognitive teaching looks like, and captured richer data on the specific area of concern.

2. Targeted professional learning was delivered. Not a generic twilight session, but three 45-minute workshops over the term, focused on the specific gaps the data revealed:

• Workshop 1: Modelling thinking processes — how to make the implicit explicit, with video examples and practice in trios
• Workshop 2: Strategy instruction — naming strategies, explaining when to use them, giving pupils practice with feedback
• Workshop 3: Structured pupil reflection — moving from "What did you learn?" to "Which strategy did you use? How do you know it worked? What would you do differently?"

Each workshop was led by the two English teachers whose observations had shown the strongest metacognitive practice — internal expertise, not an external consultant.

3. The follow-up was built into the calendar. Learning visits using the Metacognition lens were scheduled every two weeks throughout Term 2. The data would show whether the professional learning was translating into practice.

What the observation evidence showed

Term 2: 30 visits using the Metacognition lens across the same departments.

Lens effectiveness — Metacognition (Term 2 vs Term 1)

This section includes an interactive chart — open the page in a browser to explore it.

Every practice increased in frequency. But the qualifier breakdown showed whether the increase was genuine or superficial:

Qualifier shift — Term 1 vs Term 2

The strong-practice qualifiers increased substantially. The weak-practice qualifiers remained roughly flat. This meant the improvement was genuine — teachers were adding new metacognitive practices to their repertoire, not just relabelling what they were already doing.

The trend view confirmed the trajectory was sustained, not a one-off spike after the first workshop:

Monthly trend — Metacognition lens target practices

This section includes an interactive chart — open the page in a browser to explore it.

The rate of metacognition observations per visit increased each month — from 1.4 taps per visit in January to 2.3 in March — suggesting that the professional learning was embedding progressively, not just producing an initial bump.

The department picture shifted

Department comparison — Metacognition (Term 2)

The improvement was broad-based. Technologies and Modern Languages — the lowest-performing departments in Term 1 — both showed meaningful increases. The school was not relying on a couple of strong departments to pull up the average.

Connecting to HGIOS4

The observation evidence now supported a revised self-evaluation narrative for QI 2.3:

"In Term 1, structured observation evidence identified that metacognitive teaching practices were present in 45% of visits but were characterised predominantly by surface-level reflection and implicit thinking processes. Explicit strategy instruction was virtually absent. A targeted professional learning programme was implemented, led by internal expertise, and supported by fortnightly learning visits using a focused Metacognition observation lens. By Term 2, metacognitive practices were observed in 80% of visits, with strong-practice qualifiers — structured reflection, explicit strategy teaching, metacognitive prompts — increasing from 7 to 48 observations. The improvement was sustained across the term and evident across all departments."

That narrative does three things that a typical self-evaluation cannot. It quantifies the baseline. It describes the intervention with specificity. And it evidences the impact with data that an inspector can interrogate. The self-evaluation moves from assertion ("we are embedding metacognition") to evidence ("our observation data shows metacognitive strategy instruction increased from 0 to 5 observations per term, and structured reflection from 3 to 13").

The cycle continues

The Term 2 data also revealed what to focus on next. Self-monitoring — pupils using structured tools to check their own understanding during a task — remained the weakest practice in the category. "Metacognitive prompt/checklist" appeared 5 times, but "Teacher-directed monitoring only" appeared 3 times, and "Monitoring own understanding" just once. Pupils were being taught strategies but not yet deploying them independently.

That becomes the Term 3 focus. The lens narrows further. The professional learning targets the specific gap. The cycle repeats.

This is what evidence-based improvement looks like when the observation system is designed to produce structured data, not subjective recall.

For schools working with the Ofsted framework

Both HGIOS4 and the Ofsted EIF draw on the same research base — notably the EEF's Metacognition and Self-Regulated Learning guidance — so a metacognition improvement cycle is framework-neutral in substance, even where the wording differs.

The improvement-cycle structure — baseline, targeted professional learning, re-measurement — is the evidence-led practice both inspectorates look for. The lens-based data generation is what makes the cycle visible.

---

Jamie Scobie writes from extensive experience in Scottish secondary education, including pastoral care, data for improvement, and school self-evaluation. This guide is published independently: he writes in a personal capacity as the creator of Learning Lens, offering practical analysis alongside the work of building the product and reflecting on teaching, evidence, and policy. He speaks on this page only for himself and for Learning Lens, not for any employer or other organisation. He holds an MSt from Cambridge (Distinction) and a Masters from Stirling.